Zhang et al. BMC Pregnancy and Childbirth 2012, 12:119http://www.biomedcentral.com/1471-2393/12/119

RESEARCH ARTICLE Open Access

Birth weight differences between pretermstillbirths and live births: analysis ofpopulation-based studies from the U.S.and SwedenXun Zhang1, KS Joseph3, Sven Cnattingius4 and Michael S Kramer1,2*

Abstract

Background: Many stillbirths show evidence of fetal growth restriction, and most occur at preterm gestational age.The objective of this study is to compare birth weights at preterm gestational ages between live births andstillbirths, and between those occurring before or during labour.

Methods: Based on singleton births from the United States (U.S.) 2003–2005 (n=902,491) and Sweden 1992–2001(n=946,343), we compared birth weights between singleton live births and stillbirths at 24–36 completed weeks ofgestation from the U.S. and at 28–42 completed weeks from Sweden.

Results: In both the U.S. and Sweden, stillbirth weight-for-gestational-age z-scores were at least one standarddeviation lower than live birth z-scores at all preterm gestational ages (GA). In Sweden, no birth weight differencewas observed between antepartum and intrapartum stillbirths at preterm GAs, whereas birth weights amongintrapartum stillbirths were similar to those among live births at 37–42 weeks.

Conclusions: Birth weights observed at preterm gestation are abnormal, but preterm stillbirths appear to be moregrowth-restricted than preterm live birth. Similar birth weights among ante- and intrapartum preterm stillbirthssuggest serious fetal compromise before the onset of labor.

Keywords: Stillbirth, Preterm birth, Fetal growth, Intrauterine growth restriction

BackgroundDepending on jurisdiction, stillbirth is variably definedas a fetal death from 20, 22, 24 or 28 completed weeksof gestation and can also depend on birth weight [1-4].In the United States, stillbirth rates have decreased onlymodestly over the past 20 years (from 7.5 per 1,000 in1990 to 6.2 per 1,000 in 2005) [3,5,6], while infant mor-tality has fallen by more than 30% over the same timeperiod. Thus, stillbirth now accounts for over 50% of allperinatal deaths; it remains a significant and understud-ied problem [3,5]. The causes of a large fraction of

* Correspondence: michael.kramer@mcgill.ca1Departments of Pediatrics, McGill University Faculty of Medicine, Montreal,Canada2Department of Epidemiology, Biostatistics and Occupational Health, McGillUniversity Faculty of Medicine, Montreal, CanadaFull list of author information is available at the end of the article

© 2012 Zhang et al.; licensee BioMed CentralCommons Attribution License (http://creativecreproduction in any medium, provided the or

stillbirths remain unknown, even when extensive testingand autopsy are performed [7-11]. Classification ofcauses of stillbirth is often difficult and ambiguous be-cause of the “invisible” nature of the events leading todeath in utero. Many systems have been developed forclassifying causes of stillbirth, but none has been univer-sally accepted [10,11].The actual time of fetal death is usually unknown. It

has been reported that stillbirths are generally deliveredwithin 24 hours of diagnosis [12-14], and the date ofdeath for antepartum stillbirths is often arbitrarily set as2 days before delivery [15,16]. Many stillbirths show evi-dence of fetal growth restriction, and most occur at pre-term gestational ages [17-20]. However, birth weightseven among live-born preterm infants are known to belower than fetal weights of their counterparts who re-main in utero at the same gestational age [21,22]. Our

Ltd. This is an Open Access article distributed under the terms of the Creativeommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andiginal work is properly cited.

Zhang et al. BMC Pregnancy and Childbirth 2012, 12:119 Page 2 of 7http://www.biomedcentral.com/1471-2393/12/119

objective in this study was to compare the birth weightsbetween preterm stillbirths and live births, and betweenthose occurring before or during labour.

MethodsU.S. birthsWe used period linked birth-infant death data and fetaldeath data files from the National Vital Statistics Systemof the United States for the years 2003, 2004, and 2005.These publicly available data files, compiled by the U.S.National Center for Health Statistics (NCHS), includeinformation from the birth certificate on maternal socio-demographic characteristics, birth weight, period ofgestation, plurality, and live birth order and is linked toinformation from the infant death certificate. The fetaldeath files include information from all reports of fetaldeaths but do not include the cause of stillbirth [23]. Inthe United States, states vary in their definitions andreporting requirements for stillbirth [3]. Most states re-port fetal deaths at 20 weeks or more of gestation and/or 350 grams in birth weight. However, a few states re-port fetal deaths for all periods of gestation [3]. Themost recently available fetal death data file is for 2005.In the United States, gestational age (GA) is usually cal-

culated from the first day of the mother’s last menstrual

Table 1 Mean birth weight-for-gestational-age z-scores and ppreterm, and term gestation, Swedish live births and stillbirt

Mean birth weig

Live births (n=943,274)

Preterm (n=46,531) −0.29

28-31 weeks (n=4,783) −0.96

32-33 weeks (n=5,963) −0.60

34-36 weeks (n=35,785) −0.16

Term (n=899,812) +0.03

Percent of

Preterm 100 (ref)

28-31 weeks 100 (ref)

32-33 weeks 100 (ref)

34-36 weeks 100 (ref)

Term 100 (ref)

period (LMP). It has been shown that gestational agederived from the LMP estimate is prone to error, espe-cially for postterm dates [23-25]. The clinical estimate ofgestation has also been recorded since 1989. The clinicalestimate is based on the clinician’s best estimate, includingmenstrual history, physical findings, laboratory values, and(if available) sonography [24]. Recent evidence suggeststhat the clinical estimate provides rates of preterm birth,postterm birth, and GA-specific rates and relative risks ofadverse pregnancy outcomes that are more consistent withthose reported in other countries [26,27]. In this study,therefore, our analyses are based on the clinical estimateof gestational age. California does not report the clinicalestimate and was therefore excluded from the study.We restricted our primary analysis to singleton

births 24–36 completed weeks of gestation. We alsoexcluded implausibly high or low birth weights atgiven gestational ages from our study sample, basedon Table 1 in Alexander et al.[28] The proportions oflive births and stillbirths excluded were 0.6% (n=5,242)and 3.4% (n=826), respectively. Over 65% of excludedstillbirths weighted <350 g, which most states didnot report [3]. Our study sample comprised a totalof 902,491 preterm births, 23,258 (2.6%) of whichwere stillbirths.

ercent of mean live birth weight at very, moderate, latehs, 1992-2001

ht-for-gestational-age z-score P-value

Stillbirths (n=3,096)

−1.41 <0.001

Intrapartum Antepartum

−1.38 −1.41 0.249

−1.68 −1.87

−1.14 −1.41

−1.33 −1.01

−0.55 <0.001

Intrapartum Antepartum

−0.05 −0.59 <0.001

mean live birth weight

89.2 <0.001

Intrapartum Antepartum

89.0 88.9 0.283

90.8 87.9

92.3 89.3

86.2 89.6

92.7 <0.001

Intrapartum Antepartum

98.9 92.2 <0.001

Zhang et al. BMC Pregnancy and Childbirth 2012, 12:119 Page 3 of 7http://www.biomedcentral.com/1471-2393/12/119

Swedish birthsTo guard against potential misclassification of GA in theUS birth cohort [24,25], we also analyzed data from theSwedish Medical Birth Register (1992–2001). These dataare publicly available for Swedish researchers, includingone of the study authors (Professor Cnattingius). Theaccuracy of gestational age (GA), birth weight, and still-birth recorded in the Register has been previously vali-dated [29]. In Sweden, gestational age is usually based on asecond-trimester ultrasound estimate; otherwise, informationon the last menstrual period was used. In Sweden, allwomen since 1990 are offered an ultrasonic scan per-formed no later than 18 completed weeks of gestation,and 95% of the women accept this offer [4,29]. As notedearlier, the Swedish Medical Birth Register reported(until recently) only stillbirths with GA ≥28 completedweeks. Further details on the Swedish Medical BirthRegister and the study sample have been described else-where [15,30,31]. The Swedish data comprised a total of46,531 preterm births, 1,450 (3.1%) of which were still-births. The Swedish data also classify stillbirths as antepar-tum vs intrapartum, which allowed us to compare thepreterm birth weight between these two categories. We fur-ther compared birth weight between antepartum and intra-partum stillbirth at term gestation (37–42 weeks), including899,812 term births, 1,464 (0.2%) of which were stillbirths.

Construction of fetal growth standardTo evaluate fetal growth, we used estimated fetal weightas the reference standard, rather than the observed birthweight, since the latter is known to be suboptimal atpreterm GAs [21,22]. Internal population-based fetalweight standards (separate standards for U.S. and Swed-ish births) were calculated based on Hadlock’s formula[32] relating the intrauterine (ultrasound) estimated fetalweight (EFW) to gestational age in weeks: log (EFW) =0.578 + 0.332×GA – 0.00354×GA2, which predicts amean birth weight of 3619 g at 280 days. Hadlock’s for-mula assumes proportional fetal growth throughoutpregnancy and models fetal growth trajectory “by divid-ing each daily value predicted by this formula by the280-day value and fitting a third-degree polynomial ofgestational age”, which yields the following “proportion-ality equation” for GAs of 24–39 weeks [33]: % EFW =299.1 – 31.85 × GA + 1.094 × GA2 – 0.01055 × GA3.The latter equation can then be used for any estimated

280-day birth weight to estimate fetal weight in 24–39weeks of gestational age. Sex-specific birth weight z-scores for 40 weeks of gestation were first calculatedbased on the study sample, and the sex-specific weight-for-gestational-age z-scores were then extrapolated back-ward by the proportionality equation. Further details havebeen described in previous studies [15,30,31]. However,the proportionality equation does not fit well outside the

range of 24–39 weeks. In fact, the minimum value for thisequation is at 21 weeks, and thus the fetal weight pre-dicted by this formula at 20 weeks is actually higher thanthat predicted at 21 weeks. For weeks 41 and 42, observedbirth weight was used (Swedish births), since birth weightand fetal weight are similar at late GAs [30,31].

Statistical analysisThe primary outcomes were birth weight and birthweight-for-gestational-age z-score. We compared meanbirth weights and birth weight-for-gestational-age z-scores at preterm gestations between singleton stillbirthsand live births at 24–36 weeks of gestation from the U.S.and at 28–36 weeks from Sweden. We further comparedbirth weight-for-gestational-age z-scores between antepar-tum and intrapartum Swedish stillbirths both at pretermand term gestations. For all these comparisons, we usedtwo-sample t-test. At each completed week of gestation,stillbirth weight was also expressed as a percentage of themean (POM) of live birth weights at the same gestation;the mean of these percentages, with its 95% confidenceinterval (CI), was then calculated. We also carried outrace/ethnicity-specific analyses of non-Hispanic Whites,non-Hispanic Blacks, and Hispanics, from the U.S. andcompared birth weights among antepartum and intrapar-tum stillbirths from Sweden. Finally, as a sensitivity ana-lysis, we excluded from the U.S. birth cohort all birthsdiagnosed with congenital anomalies at birth and com-pared birth weight between preterm live births and still-births. All analyses were carried out using SAS version9.2 (SAS Institute Inc., Cary, North Carolina, USA).

ResultsFigure 1 shows the mean birth weight with 95% CI,Figure 2 the weight-for-gestational-age z-scores with 95%CI, and Figure 3 the percents of the mean (POM) livebirth weight for live births and stillbirths with 95% CI ateach preterm GA in the United States from 2003 to 2005.Compared to live births, stillbirth weights were lower atall GAs; the difference increased with gestational age. Onaverage, stillbirths weighed 1,013 g less than live births atpreterm gestations. Similarly, stillbirth z-scores werelower than those for live births at all preterm GAs. At24–33 weeks, live birth z-scores were slightly below 0(the reference), while at 34–36 weeks, live birth z-scoreswere slightly above 0. However, stillbirth z-scores were atleast one standard deviation lower than live birth z-scoresat all GAs between 24 and 36 weeks; the average z-scorewas −1.57, i.e., more than 1.5 SD below the reference.Moreover, stillbirth weight POMs were at least 10% lowerthan those of live births at all preterm GAs. All of thesedifferences were highly statistically significant (p<0.001).We observed similar patterns in birth weight and z-

score differences between live births and stillbirths in

500

1000

1500

2000

2500

3000

24 25 26 27 28 29 30 31 32 33 34 35 36

Gestational age in completed weeks

Gra

ms

95% upper limitLive Births95% lower limit95% upper limitStillbirth95% lower limit

Figure 1 Mean birth weight in preterm live births vs stillbirths, U.S. births 2003–2005.

Zhang et al. BMC Pregnancy and Childbirth 2012, 12:119 Page 4 of 7http://www.biomedcentral.com/1471-2393/12/119

race/ethnicity-specific analyses of U.S. non-HispanicWhites, non-Hispanic Blacks, and Hispanics (data notshown) as we did in the overall U.S. study sample. Afterexcluding all births diagnosed with congenital anomaliesat birth, we observed a similar patterns of birth weightand z-score differences between preterm live births andstillbirths (data not shown), although the average differ-ence in birth weight decreased from 1,013 to 876 g.

-2.0

-1.5

-1.0

-0.5

0.0

0.5

24 25 26 27 28 29

Gestational age

Z-s

core

Figure 2 Preterm live birth and stillbirth weight-for-gestational-age z

Table 1 compares the weight-for-gestational-age z-scores between antepartum and intrapartum stillbirthsand POMs from Sweden at very preterm (28–31 weeks),moderate preterm (32–33 weeks), late preterm (34–36weeks), and term (37–42 weeks) GAs. At preterm GAs,the differences in z-scores and POMs between antepar-tum and interpartum stillbirths were not statisticallysignificant, but (as observed in the U.S. data) the

30 31 32 33 34 35 36

in completed weeks

95% upper limitLive Births95% lower limit95% upper limitStillbirth95% lower limit

-scores, U.S. births 2003–2005.

80

90

100

24 25 26 27 28 29 30 31 32 33 34 35 36

Gestational age in completed weeks

% o

f m

ean

% of mean95% lower limit95% upper limit

Figure 3 Preterm stillbirth weights expressed as percent of the mean (POM) live birth weight, U.S. births 2003–2005.

Zhang et al. BMC Pregnancy and Childbirth 2012, 12:119 Page 5 of 7http://www.biomedcentral.com/1471-2393/12/119

differences between live births and stillbirths were highlysignificant (P < 0.001). In contrast, the z-scores andPOMs for antepartum stillbirths at term were signifi-cantly lower than for intrapartum stillbirths.

DiscussionAverage birth weights are known to be lower than intra-uterine (fetal) weights early in pregnancy [21,22],strongly suggesting that preterm births are undergrownrelative to fetuses who remain in utero at the same GA.Birth weights observed at preterm gestational ages arethus “abnormal” overall. The differences observed inSwedish births (Table 1) between preterm live birthweights and estimated intrauterine fetal weights weresimilar to those reported by Secher et al. [21] andHutcheon et al. [22]. At 32 weeks, Hutcheon et al. esti-mated that the median live birth weight is 120 g (i.e.,about 0.5 SD) lower than the median fetal weight.As shown in our study, stillbirth weights were at least

one SD below live birth weights at preterm GAs. Thefact that stillbirths weigh less at delivery than live birthsat the same GA has been previously documented[34,35]. Previous studies have also reported that amongstillbirths whose time of death is known, most are deliv-ered within 24 hours of fetal death and that the medianinterval from confirmation of cardiac activity until docu-mentation of death is 7 hours [12-14]. The fetus maylose weight in utero if the death occurs several days orweeks before delivery [3,35]. This is not likely for moststillbirths, however, since the delay between death anddelivery is often substantially lower [12-14]. The differ-ence we observed between live births and stillbirths at

preterm gestations might be due to the larger proportionof congenital anomalies in preterm stillbirths thanamong preterm live births [36]. After excluding all birthswith reported congenital anomalies, however, pretermstillbirths continued to weigh substantially less than pre-term live births.In our analysis of the Swedish data, we were surprised

not to observe significant birth weight differences be-tween antepartum and intrapartum stillbirths at pretermGAs. This result does not agree with previous reports byChard [35] and Alberman et al. [34] Alberman et al.,however, did not adjust birth weights for gestational age,and antepartum stillbirth gestational ages averaged 2weeks lower than those of intrapartum stillbirths [34],while Chard included only stillbirths with GAs 24–32weeks [35]. The absence of difference in weight betweenantepartum and intrapartum stillbirths should perhapsnot be so surprising at preterm gestations. Why wouldan intrapartum stillbirth occur at early gestation unlessit was seriously compromised before going into labour?In contrast, at GAs of 37–42 weeks, we observed a largedifference in weight between antepartum and intrapar-tum stillbirths but no difference in weight between intra-partum stillbirths and live births.Assuming little or no weight loss among most antepar-

tum stillbirths from death to delivery, preterm stillbirthsappear to be more growth-restricted than live births atthe same GA. It has been shown that size at birth isaffected by growth velocity early in gestation and thatgrowth restriction diagnosed at delivery is preceded byslower first- and early second-trimester fetal growth[37,38]. As shown in Figure 1, the difference in birth

Zhang et al. BMC Pregnancy and Childbirth 2012, 12:119 Page 6 of 7http://www.biomedcentral.com/1471-2393/12/119

weight between live births and stillbirth increasedwith advancing GA, suggesting that fetuses who sub-sequently died in utero grew more slowly before theirdeath, i.e., that fetal growth restriction in stillbirths is acumulative process.An important limitation of our study is the lack of a

good ultrasound-based fetal weight reference at pretermgestations, which would improve estimation of the dif-ference between the birth weights of live births and thefetal weights of their counterparts who remain in utero.Hadlock’s fetal growth curve [32], based on a modestnumber of White fetuses from the 1980s, does not ap-pear to fit well with our study sample (U.S. cohort), be-cause birth weight-for-gestational-age z-scores amongpreterm live births were not substantially below 0 atmost preterm GAs, as has been observed in previousstudies [21,22]. Another limitation is the lack of infor-mation on the precise timing of fetal death; we are thusunable to identify preterm fetuses who died long beforethe time of diagnosis. This is in particularly true for verypreterm stillbirths, owing to less frequent prenatal carevisits at early gestations. Furthermore, lack of informa-tion on the causes of stillbirths prevents us from explor-ing the relation between stillbirth weights and causes,and thus the relative weight deficits of stillbirths that are“unexplained.” Finally, the data used in this study arefairly old; the latest U.S. data available for fetal deathare from the 2005. There is no reason to suspect, how-ever, that the difference in birth weight between pretermlive births and stillbirths would have changed in morerecent years.Our study is further limited by the use of nation-wide

vital statistics databases, in which coding errors areknown to occur [39,40]. In particular, inaccurate estima-tion of gestational age in the U.S. has been widelyreported and discussed [24,25]. However, our sensitivityanalysis based on the Swedish data, in which gestationalage is primarily based on second-trimester ultrasoundestimates and has been well-validated [29], solidifies ourfindings. Reporting of fetal deaths is not uniform acrossthe states; a few states report fetal deaths for all periodsof gestation and for birth weight <350 g [3]. The use ofTable 1 in Alexander et al's excluded mostly extremelylow birth weight (65% of the excluded stillbirths were<350 g and nearly 3/4 were < 500 g) due to such non-uniform reporting and thus caused unbalanced exclusionbetween live births and stillbirths. Without the exclu-sion, the difference in birth weight between live birthsand stillbirths at preterm would have been larger. Theexclusion of California due to lack of data on clinical es-timate of gestation age also limits the generalizability ofour study. The proportion of births in California is about13% of the total U.S. births. Based on our previous studyof birth weight trends at term [41], in which menstrual

estimates of gestation were used in a sensitivity analysis,such an exclusion is likely to have little impact onour results.Many preterm births are growth-restricted. Our study

suggests that growth restriction is a cumulative processand that those fetuses who subsequently die preterm inutero grow even more slowly then those born alive atthe same GA. Reduced placental blood supply and theconsequent reduction in delivery of oxygen and nutri-ents to developing fetuses may both restrict fetal growthearly in gestation and increase the risk of subsequentstillbirth. Suboptimal placental function and poorgrowth early in pregnancy limit fetal growth for the re-mainder of pregnancy. Early and routine ultrasoundscans to monitor fetal size and growth during pregnancymight theoretically reduce the risk of stillbirth. This isperhaps the most important clinical implication of ourfindings. Before implementing routine clinical monitor-ing, however, an improved intrauterine fetal weightreference is required. Large-scale NIH- (U.S. NationalInstitutes of Health) and Gates Foundation-funded stud-ies of ultrasound-based fetal size assessment currentlyunder way should provide much-needed data on normalfetal growth patterns. Future research should attempt tounderstand the pathophysiological mechanisms under-lying growth restriction, as a basis for developing andtesting preventive and therapeutic interventions to re-duce stillbirth risk.

ConclusionsBirth weights observed at preterm gestation are abnor-mal, but preterm stillbirths appear to be more growth-restricted than preterm live birth. Similar birth weightsamong ante- and intrapartum preterm stillbirths suggestserious fetal compromise before the onset of labor.

Competing interestsWe have declared no conflict of interest.

Authors’ contributionsDr. XZ designed the study, performed the statistical analyses, and wrote thefirst draft of the manuscript. Drs. KSJ, SC, and MSK all contributed to thestudy design, interpretation of results, and manuscript revisions. All authorsread and approved the final manuscript.

AcknowledgementsOur study was funded by a grant from the Canadian Institutes of HealthResearch. The funding source did not contribute in any way to the design, inthe collection, analysis, and interpretation of data, nor in the writing of themanuscript and in the decision to submit the manuscript for publication.

Author details1Departments of Pediatrics, McGill University Faculty of Medicine, Montreal,Canada. 2Department of Epidemiology, Biostatistics and Occupational Health,McGill University Faculty of Medicine, Montreal, Canada. 3Department ofObstetrics & Gynaecology, Women’s Hospital and Health Centre of BritishColumbia, and School of Population and Public Health, University of BritishColumbia, Vancouver, Canada. 4Clinical Epidemiology Unit, Department ofMedicine, Karolinska Institutet, Stockholm, Sweden.

Zhang et al. BMC Pregnancy and Childbirth 2012, 12:119 Page 7 of 7http://www.biomedcentral.com/1471-2393/12/119

Received: 14 February 2012 Accepted: 25 October 2012Published: 30 October 2012

References1. Gourbin C, Masuy-Stroobant G: Registration of vital data: are live births

and stillbirths comparable all over Europe? Bull World Health Organ 1995,73:449–60.

2. Cartlidge PH, Stewart JH: Effect of changing the stillbirth definition onevaluation of perinatal mortality rates. Lancet 1995, 346:486–8.

3. MacDorman MF, Kirmeyer S: Fetal and perinatal mortality, United States,National vital statistics reports; 57(8). Hyattsville, MD: National Center forHealth Statistics; 2005:2009.

4. The Swedish Council on Technological Assessment in Health Care: Reportno. 139 [in Swedish]. Stockholm, Sweden: The Swedish Council onTechnological Assessment in Health Care; 1998:48.

5. Silver RM: Fetal death. Obstet Gynecol 2007, 109:153–167.6. Reddy UM: Prediction and prevention of recurrent stillbirth. Obstet

Gynecol 2007, 110:1151–1164.7. Collins JH: Umbilical cord accidents: human studies. Sem Perinatol 2002,

26:79–82.8. Incerpi MH, Miller DA, Samandi R, Settlage RH, Goodwin TM: Stillbirth

evaluation: what tests are needed? Am J Obstet Gynecol 1998,178:1121–1125.

9. Smith GCS, Fretts RC: Stillbirth. Lancet 2007, 370:1715–1725.10. Reddy UM, Goldenberg R, Silver R, Smith GCS, Pauli RM, Wapner RJ, et al:

Stillbirth classification - developing an international consensus forresearch. Obstet Gynecol 2009, 114:901–914.

11. Vergani P, Cozzolino S, Pozzi E, et al: Identifying the causes of stillbirth: acomparison of 4 classification systems. Am J Obstet Gynecol 2008,199:319.e1–319.e4.

12. Kellner KR, Donnelly WH, Gould SD: Parental behavior after perinataldeath: lack of predictive demographic and obstetric variables. ObstetGynecol 1984, 63:809–814.

13. Genest DR, Williams MA, Greene MF: Estimating the time of death instillborn fetuses: I. Histologic evaluation of fetal organs; an autopsystudy of 150 stillborns. Obstet Gynecol 1992, 80:575–584.

14. Radestad I, Steineck G, Nordin C, Sjogren B: Psychological complicationsafter stillbirth - influence of memories and immediate management:population based study. Br Med J 1996, 312:1505–1508.

15. Clausson B, Gardosi J, Francis A, Cnattingius S: Perinatal outcome in SGAbirths defined by customised versus population-based birthweightstandards. Br J Obstet Gynaecol 2001, 108:830–834.

16. Zhang X, Platt RW, Cnattingius S, Joseph KS, Kramer MS: Are babies born toshort, primiparous, or thin mothers “normally” or “abnormally” small?J Pediatr 2007, 150:603–607.

17. Cnattingius S, Haglund B, Kramer MS: Differences in late fetal death ratesin association with determinants of small for gestational age fetuses:population based cohort study. Br Med J 1998, 316:1483–1487.

18. Gardosi J, Mul T, Mongelli M, Fagan D: Analysis of birthweight andgestational age in antepartum stillbirths. Br J Obstet Gynaecol 1998,105:524–530.

19. Gardosi J, Kady SM, McGeown P, Francis A, Yonks A: Classification ofstillbirth by relevant condition at death (ReCoDe): population basedcohort study. Br Med J 2005, 331:1113–1117.

20. Froen JF, Gardosi J, Thurmann A, Francis A, Stray-Pedersen B: Restrictedfetal growth in sudden intrauterine unexplained death. Acta ObstetGynecol Scand 2004, 83:801–807.

21. Secher N, Hansen P, Thomsen B, Keiding N: Growth retardation in preterminfants. Br J Obstet Gynaecol 1987, 94:115–120.

22. Hutcheon JA, Platt RW: The missing data problem in birth weightpercentiles and thresholds for "small-for-gestational-age”. Am J Epidemiol2008, 167:786–792.

23. Vital Statistics of the United States: 2003 Natality. Hyattsville, MD: NationalCenter for Health Statistics; 2005.

24. Klebanoff MA: Gestational age: not always what it seems. Obstet Gynecol2007, 109:798–799.

25. Martin JA: United States vital statistics and the measurement ofgestational age. Paediatr Perinat Epidemiol 2007, 21:13–21.

26. Ananth CV: Menstrual versus clinical estimate of gestational age datingin the United States: temporal trends and variability in indices ofperinatal outcomes. Paediatr Perinat Epidemiol 2007, 21:22–30.

27. Joseph KS, Huang L, Liu S, Ananth CV, Allen AC, Sauve R, et al: Reconcilingthe high rates of preterm and postterm birth in the United States. ObstetGynecol 2007, 109:813–822.

28. Alexander GR, Himes JH, Kaufman RB, Mor J, Kogan M: A United Statesnational reference for fetal growth. Obstet Gynecol 1996, 87:163–168.

29. Cnattingius S, Ericson A, Gunnarskog J, Kallen B: A quality study of amedical birth registry. Scand J Soc Med 1990, 118:143–148.

30. Zhang X, Platt RW, Cnattingius S, Joseph KS, Kramer MS: The use ofcustomized vs population-based birth weight standards in predictingperinatal mortality. Br J Obstet Gynaecol 2007, 114:474–477.

31. Hutcheon JA, Zhang X, Cnattingius S, Kramer MS, Platt RW: Customisedbirthweight percentiles: does adjusting for maternal characteristicsmatter? Br J Obstet Gynaecol 2008, 115:1397–1404.

32. Hadlock FP, Harrist RB, Martinez-Poyer J: In-utero analysis of fetal growth: asonographic weight standard. Radiology 1991, 181:129–133.

33. Gardosi J, Mongelli M, Wilcox M, Chang A: An adjustable fetal weightstandard. Ultrasound Obstet Gynecol 1995, 6:168–174.

34. Alberman E, Blatchley N, Botting B, Schuman J, Dunn A: Medical causes onstillbirth certificates in England and Wales: distribution and results ofhierarchical classifications tested by the Office for National Statistics. Br JObstet Gynaecol 1997, 104:1043–1049.

35. Chard T: Does the fetus lose weight in utero following fetal death: astudy in preterm infants. Br J Obstet Gynaecol 2001, 108:1113–1115.

36. Liu S, Joseph KS, Wen SW, Kramer MS, Marcoux S, Ohlsson A, et al: Seculartrends in congenital anomaly-related fetal and infant mortality inCanada, 1985–1996. Am J Med Genet 2001, 104:7–13.

37. Milani S, Bossi A, Bertino E, Di Battista E, Coscia A, Aicardi G, et al:Differences in size at birth are determined by differences in growthvelocity during early prenatal life. Pediatr Res 2005, 57:205–210.

38. Smith GC, Smith MF, McNay MB, Fleming JE: First-trimester growth andthe risk of low birth weight. N Eng J Med 1998, 339:1817–1822.

39. Schoendorf KC, Branum AM: The use of United States vital statistics inperinatal and obstetric research. Am J Obstet Gynecol 2006, 194:911–915.

40. Northam S, Knapp TR: The reliability and validity of birth certificates.J Obstet Gynecol Neonat Nur 2006, 35:3–12.

41. Zhang X, Joseph KS, Kramer MS: Decreased term and posttermbirthweight in the United States: impact of labour induction. Am J ObstetGynecol 2010, 203:124.e1–124.e7.

doi:10.1186/1471-2393-12-119Cite this article as: Zhang et al.: Birth weight differences betweenpreterm stillbirths and live births: analysis ofpopulation-based studies from the U.S. and Sweden. BMC Pregnancy andChildbirth 2012 12:119.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit